Output Explorer

Every prompt in the paper, and what each model wrote back.

Answer a question about a scanned page from its OCR text. Scored by exact match and ANLS against the accepted answers.

13 of 5,330 prompts

Nearby prompts. All 5,330 DocVQA prompts

PromptDocVQA · pagenzwm0227_3.png

Who comes under BOARD OF TRUSTEES, according to the flowchart?

OCR text of the page · 382 characters; the scanned image itself is not published
THE INTERNATIONAL LIFE SCIENCES INSTITUTE
BOARD OF
BOARD OF
TRUSTEES
MEMBERS
OFFICERS
OF THE
INSTITUTE
INDUSTRY TECHNICAL COMMITTEES
SPECIAL
ACTIVITIES
SCIENTIFIC EXPERT COMMITTEES
FOOD
HANOVER
CAFFEINE
COLOR
GUIDELINE
PATHOLOGY/
NUTRITION
USERS
MEETING
TOXICOLOGY
ORAL
SUBCOMMITTEE ON
SUBCOMMITTEE ON
HEALTH
PHOSPHATES
STATISTICS
TERATOLOGY
- -
- -
SUBCOMMITTEE ON
1 -
EPIDEMIOLOGY
System prompt · identical for every setup
Answer the question using only the OCR text from a single document page. Return only the answer, with no explanation. Preserve the answer wording from the OCR text when possible.
Expected answer
OFFICERS OF THE INSTITUTEOfficers of the institute
Models
4 of 4 columns · click a model to add or remove it

Ours

Wrong

MEMBERS, OFFICERS, OF THE INSTITUTE, INDUSTRY TECHNICAL COMMITTEES, SPECIAL ACTIVITIES, SCIENTIFIC EXPERT COMMITTEES

116 characters35 tokens

Aux 2015

Wrong

BOARD OF TRUSTEES

17 characters7 tokens

PiT-FT 2015

Wrong

Empty response.

0 characters

ChronoGPT 2015

Wrong

BOARD OF TRUSTEES BOARD OF TRUSTEES MEMBERS OFFICERS OF THE INSTITUTE INDUSTRY TECHNICAL COMMITTEE SPECIAL ACTIVITIES SCIENTIFIC EXPERT COMMITTEE FOOD HAN

154 characters64 tokens